Document Abstraction Engine for Legal Entity Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Leases and mortgages often contain complex and archaic language, making it difficult for individuals to identify key components within these documents.
Innovation Solution
A document abstraction engine that utilizes a combination of natural language processing and machine learning techniques to analyze legal documents, extract key entities, and generate a summary by arranging these entities according to an ontology dedicated to the document type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional manual analysis of legal documents is used, then comprehensive understanding of document content is achieved, but time consumption and accessibility difficulty increase
Solution Approach 1:
The patent replaces manual mechanical analysis of legal documents with an automated AI-based system. The system uses transformer models and named entity recognition to automatically extract key information, replacing the manual reading and analysis process. This substitution eliminates the time consumption associated with manual document review while maintaining comprehensive information extraction.
Solution Approach 2:
The system enables self-service document analysis where the AI model autonomously processes legal documents without human intervention. The transformer models automatically identify entities, relationships, and key components, allowing users to obtain document summaries and key information extraction services automatically, eliminating the need for manual professional analysis.
2Adaptability or versatility
If complex archaic language is used in legal documents, then comprehensive legal coverage is achieved, but ease of understanding and identification of key components decreases
Solution Approach 1:
The patent introduces an intermediary AI system that bridges the complex archaic legal language and user understanding. The transformer models act as intermediaries that interpret the complex legal terminology and translate it into structured, easily understandable summaries. This intermediary layer preserves the comprehensive legal coverage while making the content accessible to users without legal expertise.
Solution Approach 2:
The system extracts only the essential key components and entities from the complex legal document text. By using named entity recognition and transformer models, the system identifies and extracts critical information such as parties involved, dates, amounts, and legal obligations, separating these from the surrounding archaic language. This extraction provides comprehensive legal coverage while presenting information in an easily understandable format.
3Measurement precision
If multiple named entity recognition models are used, then extraction precision of entities is improved, but system complexity increases
Solution Approach 1:
The patent merges multiple named entity recognition models into a unified system architecture. Instead of using separate independent models, the system combines transformer models and other NER models into an integrated pipeline where models work cooperatively. This merging approach maintains high entity extraction precision through multiple models while reducing overall system complexity by sharing computational resources and coordinating processing.
Data Source
AI summary
A computing system receives a document to be analyzed. The document is associated with a document type of a plurality of document types. The computing system determines the document type associated with the document. The computing system routes the document to a plurality of name entity recognition transformer models trained to identify a plurality of entities in the document. The plurality of name entity recognition transformer models extracts the plurality of entities from the document. For each word in the document, a multi-modal encoder-based transformer model determines a probability that the word is an entity based on output generated by the plurality of name entity recognition transformer models and the document. The computing system generates a summary of the document by arranging the plurality of entities in accordance with an ontology dedicated to the document type.